Students from under-served and minority communities are not only under-represented in the health professions but also academically unprepared to succeed in majors leading to those jobs. Existing applications have been unsuccessful in raising math scores, in large part because two-thirds of educational technology licenses purchased by schools are never used. To be usable in educational programs, software must accommodate limits on budget, hardware and Internet yet still offer engagement of students, ease of installation and use. Our goal is to develop and test a prototype integrated software platform to rapidly create mobile game-based learning software that is engaging, effective and runs on low-end devices. The innovation in Empiric Empires is four-fold; 1) a platform that develops software faster and cheaper, 2) emphasis on character and narrative, over video and 3D graphics, enabling installation on low-end mobile devices 3) educational and assessment content embedded in a true adventure game and 4) cross-curricular content including mathematics, social studies and health science. In-game data collection of players? answers to math and health science questions provides a baseline of student pre-existing knowledge and a measure of effectiveness of the instructional material. Video is largely replaced with voiceovers and animated gifs, graphic novels and ?text-messaging? between characters. We will validate this platform with the creation of Aztec Era, a mobile app for middle school mathematics standards with incorporation of concepts from epidemiology and biostatistics. We will conduct usability testing on Aztec Era with a maximum variation sample of 60 middle school students. Testers will be drawn from programs in three sites: a diverse urban community in California and American Indian reservations in North and South Dakota. Qualitative data generated from interviews and observations will be summarized in five user case studies, identifying typical and ?edge? users. Throughout the project, quantitative data will be collected electronically on frequency and duration of gameplay sessions, modules completed and correct or incorrect responses to STEM challenges. Descriptive statistics will be generated for number of learning tasks completed, correct answers, for mathematics and health science questions, overall and by site. MANOVA will be used to test for differences by gender and site. The prototype phase ends with a completed design document, data on usability and a codebase of data collection activities, game-based instruction and in-app reinforcers that can be expanded in Phase II development of a commercial product. Having solved the problem of getting students to use the software regularly in Phase I, we can assess the impact of use on knowledge of mathematics and public health science and interest in health careers in Phase II. !